Punishment's Legal Templates: A Theory of Formal Penal Change

Abstract

The well-known gap between law on the books and law in action often casts doubt on the significance of changes to law on the books. For example, the rise and fall of penal technologies have long been considered significant indicators of penal change in socio-historical analyses of punishment. Recent research, however, has challenged the significance of apparently large-scale penal change of this kind. This article clarifies the significance of penal technologies' rise and fall by offering an alternative account of formal penal change, introducing the analytical concept of “legal templates,” structural models of legal activity (e.g., punishment) available for authorization and replication across multiple jurisdictions. Analyzing punishment's templates explains how new penal technologies can be important harbingers of change, even when they fail to revolutionize penal practice and are not caused by a widespread ideological shift. This article locates the significance of punishment's legal templates in their constitutive power—their ability, over the long term, to shape cognitive-cultural expectations about what punishment is or should be. This power appears only when the template is widely adopted by a plurality of jurisdictions, thereby becoming institutionalized. Ultimately, these institutionalized templates define the scope of future punishment.

A wide array of law and society scholarship has described the law as a severely limited social institution with uncertain influence. Examples of the law's limits abound, from the many unlawfully injured people who, prevented by their vulnerable status, do not mobilize the law (Albiston 2005; Bumiller 1987; Nielsen 2000), to disputants who prefer business and community norms over law (Ellickson 1986; Macaulay 1963). As a tool of social change, the law frequently disappoints: major court victories and much-celebrated legislation have been limited in their effect (Edelman 1992; Rosenberg 1991). In myriad accounts, legal change of many stripes falls short of expectations because there is a significant, well-examined gap between law on the books and law in action (Friedman 2016).

These findings pose a dilemma: in what sense do changes to law on the books matter? When official policy rhetoric and actual practice diverge extensively, what is the significance of different rhetoric and new laws? One version of this dilemma appears in differing accounts of legally authorized changes to the available forms of punishment, or formal penal change: while some accounts describe the adoption of new penal technologies¹ as signaling important sea changes in penality, other accounts caution that such shifts hide greater continuities across time and heterogeneity across place. What, then, is the significance of formal penal change, particularly when its impact on practice is smaller, different, or more variable than intended or expected?

This article uses formal penal change as a case study to clarify the significance of legal change. Drawing on neo-institutional theory from organizational sociology (DiMaggio and Powell 1983; Edelman 1992, 2016; Meyer and Rowan 1977; Tolbert and Zucker 1983), it proposes a theory of formal penal change. This theory begins by deconstructing punishment into its constituent legal templates, or individual structural models of punishment available for authorization and replication across multiple jurisdictions, whether defined as legal, political, or organizational units.² Although extant studies of formal penal change have focused on penal technologies, this study shifts the focus to their underlying legal templates, thereby splitting penal technologies into their intention—the guiding idea or principal organizing structure (the template)—and practice—how they play out in reality. Moreover, as the formal structures of punishment, templates are distinct from other outcomes of interest, such as individual-level decisions to punish, trends in punishment rates, and the popularity of certain logics or justifications of punishment.

Next, I repurpose insights from neo-institutional theory and the law and society literature to construct a generalizable narrative of formal penal change, emphasizing legal templates' diffusion. First, many jurisdictions will adopt a legal template not for its effectiveness at crime control, but because it has been institutionalized. Institutionalized legal templates are legitimacy-granting structures: formally adopting these templates gives a jurisdiction legitimacy. Second, institutionalized legal templates may have limited impact on actual practice when jurisdictions adopt (or retain) the template simply to emphasize its visible conformity to normative expectations; such jurisdictions can privately pursue their technical goals by loosely coupling the template to actual practice. Finally, institutionalized legal templates can change official and lay understandings of what constitutes acceptable punishments. Even in the absence of meaningful change on the ground, such legal templates can shape long-term expectations about punishment and the legitimacy of jurisdictions, organizations, and actors meting out punishment.

By examining templates separately from their impact on actual practice, the template's diffusion becomes one of the most important, but under-emphasized, analytical sites for understanding the significance of formal penal change. Whereas most accounts of formal penal change focus on factors driving templates' innovation or initial adoption, this theory focuses on the reasons for, and consequences of, templates' diffusion. This analytic shift from innovation to diffusion changes how we envision penal change over time—from a linear timeline to a messy, variegated, incomplete process—in a way that complicates standard theories about the engines, mechanisms, and significance of penal change. In doing so, it moves beyond earlier explanations—for example, that penal change in the form of new penal technologies will have a particular impact on practice, serving as the physical manifestation of new values that brought about the new technology—that have been recently discredited (Goodman et al. 2017). Ultimately, this article theorizes the conditions under which formal penal change matters, even when its impact on practice is limited or different from expectations.

Penal change is only one example of legal change (Lynch 2011), but this theory is generalizable to other areas of law. Indeed, we can think of changes to law on the books—the rise and fall of new rights, jury trials, employment regulations, constitutions, problem-solving courts, alternative dispute resolution fora—as legal templates, or particular models of legal activity available for authorization and replication. Applying this theory encourages a broader, more historical analysis that examines these developments not in isolation but as part of a longer sequence of developments. This longer look reveals templates' broader significance: a legal template may facilitate only limited changes to law in action but, if institutionalized, it may have lasting power as constitutive of future expectations about the law.

That the law is constitutive of reality is certainly not a new insight to the law and society field (e.g., Ewick and Silbey 1998; Gordon 1984; Hirsch and Lazarus-Black 1994; Starr and Collier 1989). Instead, building on the neo-institutional tradition within law and society and its attention to the law's constitutive power (Suchman and Edelman 1996), this article seeks to think through (and historicize) the process of institutionalization and how the elements of that process enable the law to become constitutive. Indeed, as I explain later, not all templates become constitutive—they must be institutionalized to have that power. Additionally, the template itself need not change the practical reality—practices on the ground, how things are really done—to change the way people think about how things should be done.

Theorizing Penal Change

Defining Penal Change

When discussing penal change, scholars mean change in the way a society or jurisdiction (defined locally, nationally, or otherwise) treats its perceived and convicted criminals. Most concretely, penal change refers to patterned, widespread differences, or trends, emerging in the daily practice of punishment—those individual decisions that become visible when aggregated, such as decisions to arrest, convict, divert to probation, sentence to lengthy prison sentences, or early release from prison on parole. Important examples include mass supervision, prison overcrowding, and sentencing disparities.

Changes in penal trends, however, are actually the product of what I will label formal and informal penal changes.³ Formal changes are legally authorized or official changes in a jurisdiction's available forms of punishment: the authorization of the first prisons, the abolition of capital punishment, the rise of a particular style of or approach to policing, the passage of particular sentencing laws. These changes are immediately visible on the penal landscape as changes to law on the books (whether literally in statute books, through case law, or on agency websites and manuals). Before they directly affect those individuals eligible for punishment, however, such changes must be implemented by those who directly control the punishment experience.

Informal changes include changes in the values, norms, biases, or beliefs about crime, criminals, and punishment, such as beliefs that crime is on the rise, criminals are inherently sinful and irredeemable, extant punishment is too severe, punishment should rehabilitate or incapacitate, punishment should be cheap or even profitable, or punishment should reinforce social (class, race, gender) hierarchies. These informal changes may manifest in media coverage, public opinion polls, or the organizational culture of criminal justice agencies (e.g., courts, police departments), but they only lead to penal trends by influencing the behavior of jurors, judges, prosecutors, police officers, parole board members, lawmakers, voters, and other decision makers.

Both formal and informal changes are important for understanding changes in penal trends, and scholars frequently examine both types of changes simultaneously. However, informal and formal penal change contribute to penal trends differently. Formal changes structure individual decisions by changing the range of options from which one can select; informal changes structure those decisions by providing guidance about which option one should select. For example, a judge must adhere to the law's range of sentencing options, but her own values and her county's socio-political context may encourage her to select one of the more severe options. Likewise, agency rules may require a parole officer to revoke parolees for dirty urine samples, but his own view of the rules, past experience, and sense of morality may lead him occasionally to ignore such infractions.

Thus, by distinguishing between informal and formal changes, we can better understand the exact causal pathway of changes in penal trends. This article is primarily concerned with formal penal change and its consequences.

Problematizing Theories of Formal Penal Change

Macro-level social theoretical explanations of punishment and penal trends have traditionally focused on formal changes, especially the emergence and decline of penal technologies. These technologies may include a new method or implement of punishment (e.g., incarceration, execution, community supervision) or a new subtype of an existing method (e.g., the supermax prison, the single-drug protocol for lethal injection, intensive supervision probation). Most famously, Durkheim (1893), Rusche and Kirchheimer (1939), and Foucault (1977) viewed the rise of modern state prisons, roughly coterminous with capital and corporal punishments' decline, as a significant dividing line in the history of punishment. Indeed, the rise or fall of such penal technologies have usefully served as temporal boundaries demarcating penal eras.

A more fragmented version of this approach has continued with recent analyses of the rise of mass incarceration and the related penal complex that created and sustains it. Thus, scholars have examined the appearance of the warehouse prison (Irwin 2005), the supermaximum security (supermax) prison (Reiter 2016), three strikes laws (Karch and Cravens 2014; Zimring et al. 2001), sex offender laws (Leon 2011), and drug-testing policies for parolees and probationers (Feeley and Simon 1992) as significant arrivals on the penal landscape. All of these new technologies have been adopted widely, seem to epitomize the punitive turn, and have either contributed to the growth in incarceration rates or qualitatively changed the experience of punishment; they are thus deemed important causes and examples of penal change.

As numerous scholars have noted, however, seemingly widespread formal changes may not be as significant as they appear. Goodman et al. (2017) have recently summarized three problems facing accounts about sudden, significant changes in punishment. First, such accounts overstate the homogeneity of any penal era. While penal trends may be apparent nationwide, certain states or regions may shun dominant trends or only adopt these trends reluctantly, creating substantial variation across time and place (Lynch 2011). A growing number of historical case studies of individual states—Illinois (Jacobs 1977), Florida (Schoenfeld 2014), Pennsylvania (Rubin 2013, 2020), Arizona (Lynch 2010)—have rejected nationwide trends, often holding out until the trend had expired in most other states. Moreover, the politics of punishment that helped precipitate new sentencing laws vary by state, leading to differences in the levels of punitiveness and discretion embedded in such laws (Barker 2009). For example, California's three strikes law is notoriously extreme relative to the versions of the law adopted by most other states, affecting far more people than in any other state (Zimring et al. 2001). Traditional accounts overlook the spatial and temporal heterogeneity in otherwise national trends.

Second, these accounts misstate the level of support for apparently dominant penal trends; instead of reflecting a new consensus over the goals and methods of punishment, penal change is the result of conflicting and changing power dynamics that enable some groups to influence policy in ways they had not before (Goodman et al. 2017). Seemingly popular changes, even those voted in by an enthusiastic public, are often the result of a small but powerful minority. Penal policies of the last 40 years, for example, have largely reflected the values of suburban or rural voters rather than the large but less powerful contingent of urban voters (Miller 2008). Indeed, lobbyists and interest groups have been essential in facilitating penal change (Gottschalk 2006). A 2004 voter initiative was poised to mitigate the most extreme features of California's notorious three strikes law until advertisements funded by the state's powerful prison guard union and victims' rights lobby reversed public opinion on the law (Page 2011). Even among state actors, we find disagreements over penal policy within the ranks of the corrections staff (Goodman et al. 2017) and between correctional administrators, the governor, and the legislature (Feeley and Rubin 2000; Lynch 2010; Rubin and Phelps 2017; Schoenfeld 2011). Traditional accounts of penal change overlook this subdermal contention about ostensibly dominant policies and technologies.

Finally, these accounts overstate the reality of penal experience: in any given locale, large and important gaps frequently emerge between policy discourse and the implementation of official mandates. Much recent penal change has happened at the level of rhetoric, albeit enshrined in law, such as the rejection of a rehabilitative approach to punishment (Allen 1981) in favor of a risk-management approach (Feeley and Simon 1992). However, actual spending on treatment programs and staff remained relatively robust for decades after penal rhetoric hardened (Phelps 2011) while frontline workers rejected the risk-management orientation of many new parole and probation policies in favor of investigatory approaches (Lynch 1998). Across prison facilities, moreover, variation in the extent to which correctional personnel agree with the new ideology provides different experiences of punishment to prisoners within the same state (Kruttschnitt and Gartner 2005). Traditional accounts overlook the significant gap between the reigning theory or official representation of punishment and on-the-ground practice.

For these reasons, Goodman et al. (2017) argue, apparently major shifts in penal policy, particularly those signaled by formal changes, are not only misleading, but manifest at the wrong level of analysis to reveal what causes penal change. These limitations suggest that current tools for discussing and analyzing penal change—by overstating geographical and temporal homogeneity, implying broader support, and focusing on externally visible changes—prevent us from fully understanding the dynamics of change. For Goodman et al., these limitations invalidate the previously dominant approach to describing the history of American penal change as a constant pendular movement between distinct periods of rehabilitative and punitive approaches to punishment demarcated by the arrival of new technologies (e.g., the prison, parole) and different types of technologies (the penitentiary, the factory-style prison, the plantation-style prison, the Big House, the correctional institution, and the warehouse prison) (e.g., Irwin 2005). More than “breaking the pendulum [metaphor],” however, these limitations also cast doubt on the significance of new penal technologies, particularly as harbingers of new eras in punishment.

And yet, some scholars retain an almost instinctual commitment to the emergence of new technologies as significant indicators of change. Indeed, for Garland (2003: 64), the recent punitive turn does not signify a new, distinct penal era because there are no “new institutions and practices being legislated into existence,” only changes to penality's “objectives and orientation.” Ultimately, he argued, “Temporary and reversible conjunctural shifts must be distinguished from long-term structural transformations” (Garland 2003: 68, emphasis added). By this metric, modern penality, distinct from what came before, was forged in the decades around 1900, with the emergence of parole, probation, and other new penal technologies (Garland 1985: 5). The heightened use of these technologies simply represents a later stage of the same era—“high” or “late” modernity—rather than the beginning of a new era—“post modernity” (Garland 2003: 61). While this approach provides a conservative definition of change (e.g., overlooking new versions of older technologies), it conveys the lasting centrality of new penal technologies in analyses of penal change. However, in light of recent findings, scholars' commitment to penal technologies as significant phenomena—especially markers of penal change—needs to be defended.

In the remainder of this article, I propose a theory of formal penal change that rebuilds the significance of penal technologies. Specifically, the theory keeps penal technologies at the center of analyses while overcoming the three limitations associated with accounts that assume the significance of penal technologies' rise and fall. I begin with the assumption, drawing on Goodman et al. and contra traditional accounts, that when new penal technologies emerge, they do not reflect widespread beliefs about crime and punishment; rather, they reflect beliefs popular among or promoted by some locally specific, influential group (e.g., social elites). Consequently, rather than exploring the causes of technologies' innovation, I emphasize the stages after innovation, focusing on the process by which a given technology becomes dominant.

In the next section, I shift our focus from penal technologies to legal templates. Building on core insights from neo-institutional theory and law and society, I emphasize that, when adopted for symbolic purposes, legal templates are not faithfully applied in practice as intended by their innovators, but are used by local authorities in unanticipated ways. Therefore, in the subsequent section, I demonstrate how legal templates' diffusion determines their significance irrespective of their anticipated impact on actual practice. Next, I lay out three main parts of the legal templates theory. Although I use diverse examples throughout to illustrate the theory's broad applicability, in the penultimate section, I systematically apply the theory to two overlapping periods of penal change. This legal templates theory provides a more precise, alternative understanding of penal change and penal practice following the emergence of new penal technologies that rise to prominence.

Legal Templates of Punishment

Definition

Legal templates of punishment are idealized, replicable models that specify, in varying degrees of detail, punishment's structure—what punishment should look like, of what activities it would consist, who is involved, where it takes place, or how long it lasts. As descriptions or models that can be legally authorized or formally adopted, templates' authorization often appear in statutes and court cases, but their histories need not begin there: reformers or penal administrators may first create a template in their writings or routines that only later receives formal authorization.

Importantly, templates are not justifications of punishment, and they are conceptually distinct from logics, or the “organizing principles” that motivate action (Friedland and Alford 1991: 248). Penal logics, whether traditional (e.g., rehabilitation, deterrence) or new (e.g., harm reduction, risk management), identify a particular goal for punishment and provide a general account of how punishment will achieve that goal. By contrast, legal templates describe the particular forms punishment will take, whatever the underlying logic. Indeed, templates can simultaneously be justified by multiple logics and their justifying logics may change over time.

Instead, legal templates can be understood as the “formal structures” of punishment. Formal structures are “a blueprint for activity” with “elements [that] are linked by explicit goals and policies that make up a rational theory of how, and to what end, activities are to be fitted together” (Meyer and Rowan 1977: 342). As formal structures, legal templates of punishment are the official account or representation of the work to be done by those who mete out punishment; they “often serve only as a decorative facade concealing the ‘real’ agenda and structure” (Scott 2003: 28). I use the label “template,” rather than formal structures, to highlight two characteristics: First, as models, templates are available to be borrowed, copied, and authorized by multiple jurisdictions. Second, where a template has been authorized or adopted, its relationship to actual practice will vary across time and place. More than a definition, the templates concept is used to invoke a particular analytical approach, developed more below, that is flexible enough to describe a range of penal phenomena, much of which has previously been referred to as “penal technologies.”

In many contexts, the term legal templates can replace the terms “penal technologies” or “penal techniques” (albeit stripped of their Foucaultian undertones). These terms are not standardized in the literature: “technology” is sometimes used interchangeably with “technique” and “institution,” each of which means different things to different scholars, thus rendering the terminology confusing and less useful. This article offers a clearly defined alternative vocabulary. Importantly, the term “template” has an advantage over “technologies” and “techniques”: when scholars describe the adoption of a penal technology, they may be speaking about either the model (or description) of punishment or its actual use. The term “template” emphasizes the model and not its actual use.

Different modes of punishment and social control—execution, incarceration, community supervision, fines, policing, and so forth—each have their own sets of templates. Executions, for example, have been guided by different templates in different times and places across American history: hanging, firing squad, electric chair, gas chamber, and lethal injection. Some templates can be subdivided into distinct versions: The hanging template includes both strangulation-style hangings and drop-style hangings; likewise, the lethal injection template includes both a three-drug protocol and a single-drug protocol. I call these different styles or protocols “versions” because they are substantially similar forms of the same execution method or template; by contrast, the electric chair and gas chamber are templates as they are substantially different methods of execution. A jurisdiction is unlikely to rely on different versions of the same template simultaneously (e.g., states rarely authorize both single- and triple-drug protocols for lethal injection); but a jurisdiction may rely on multiple templates for a single mode of punishment (e.g., many states authorize both lethal injection and some other template for execution such as hanging, firing squad, electrocution) (see Banner 2002; Garland 2010). Considering that each jurisdiction frequently employs multiple modes of punishment simultaneously (from execution to community supervision), each jurisdiction will often have multiple templates at any time. Table 1 offers a selected list of the various templates that have characterized execution, incarceration, and community supervision throughout American history.

Table 1. Legal Templates of Punishment from American History. Select versions of various templates are included as illustrations.

While legal templates have permeated penal systems throughout history, the most readily identifiable templates are typically the most popular, widely adopted templates. Many of these templates, because of their widespread adoption, became iconic representations of punishment and have received disproportionate scholarly attention.⁴ Such popularity, however, is not definitional to the template; execution via firing squad was never widely adopted, but it is a template nonetheless as it offers a more or less specific vision of execution that can therefore be authorized. However, as the rest of this article will explain, only those templates that are copied can achieve long-term significance.

Templates on the Ground

Definitionally, actual practice—punishment as meted out “on the ground”— is distinct from the template that guides it. Templates are idealized models—putative descriptions of, or prescriptions for, punishment; consequently, templates often diverge from their on-the-ground implementation. Consequently, even our most distinctive templates show much greater variation in practice than their categorical descriptions suggest. To take one template as an example, the supermax prison is often defined as a facility containing a prison system's most dangerous prisoners in solitary confinement in conditions of sensory-deprivation for 23 or more hours a day. In practice, however, it is difficult to create an inventory of the nation's supermaxes because of their different names and varied internal practices (Reiter 2012a). Moreover, some states relying on recognizable and even prototypical supermaxes double-cell their prisoners, apparently violating the very essence of the supermax (solitary confinement) and the rationale for prisoners' confinement (prisoners considered too dangerous to be around other prisoners are confined together) (Reiter 2012b). In this and other cases, the template's description bears a tenuous connection to reality.

Legal templates are not, however, wholly irrelevant to the daily practice of punishment. Even when implemented imperfectly, templates can significantly impact accused and convicted criminals' lives. Indeed, nationwide, thousands of supermax prisoners spend months, years, or decades in solitary confinement (Reiter 2016). However, legal templates' effect on actual practice frequently differs significantly from their creators or promoters' expectations, whether progressive or repressive (see Cohen 1979). For example, correctional administrators in the 1980s designed supermax prisons for the “worst of the worst”; in practice, correctional officers use their extensive discretion to send myriad prisoners to supermax confinement even for minor rule violations (Reiter 2012c: 543; Reiter 2016). Moreover, supermax prisons were supposed to be used on a select number of prisoners for short periods preceding a step-down period whereby prisoners would gradually reacclimatize to human contact. Instead, supermaxes contain so many prisoners for such long periods that the facilities have become overcrowded, while subsequent administrators never implemented a step-down period (Reiter 2012c: 540). Thus, legal templates do affect actual practice, but their impact is often partial, unanticipated, or marginal to the template's design.

It is this divergence between practice and description that has most called into question the significance of penal change associated with (what I am calling) legal templates. The primary challenge, then, becomes identifying the significance of the legal template divorced from its immediate impact on punishment in practice. As we shall see, this reframing improves our understanding of penal change by changing how we think about, envision, and describe formal penal change.

Focusing on Templates' Diffusion

Legal templates adopted by a single jurisdiction alone are insignificant in macro-level analyses of penal change. Rather, to be significant as a social phenomenon at this level, a template must be replicated on a large scale: it must be adopted widely, across some discrete set of jurisdictions, such as the American South, the European Union, or the Global South.⁵ Indeed, scholars pay little attention to the handful of American states that retain the firing squad or hanging as optional methods of execution along with the more common lethal injection because minor trends are not as analytically intriguing as major trends. Likewise, Arizona and other sunbelt states were outliers or laggards for much of the twentieth century for their regionally distinctive “sunbelt justice” model of tough and cheap punishment. They were not penologically significant states until the sunbelt, and Arizona in particular, became the bellwether for late-twentieth-century American punishment, developing multiple templates adopted around the country.⁶ It is, I argue, this widespread adoption, or diffusion, that signals major penal change.

To date, penal scholars have neglected diffusion in favor of innovation; but focusing on the “moment” of innovation has distorted our understanding of penal change (Rubin 2015). Scholars of American prison history, for example, explicitly or implicitly assume a somewhat staccato vision of penal change that generalizes from influential, innovative states. Prison history is typically divided into discrete periods (usually two or three decades) associated with the rise of a type of prison and related technologies (see the lower portion of Figure 1) (e.g., Irwin 2005). This approach has been used on a larger timescale (sometimes centuries instead of decades) at the international level (e.g., Simon 2013).⁷ Thus far, it has been a useful heuristic device for clarifying apparently significant changes across large swaths of time and space. However, following the recent literature, it may be time for an alternative approach.

Figure 1. Comparing the traditional and legal templates approaches to prison history, 1790s to 1930s. The arcs above the line represent the diffusion of particular legal templates over time. An arc's height and width represent the number of jurisdictions that adopted the template: the beginning and end of an arc (width) represent the first and last adoptions of the template while the highest point of the arc roughly corresponds to the time that the template has been adopted by most of the jurisdictions that will adopt it. The overlap across arcs illustrates the degree to which the end of one template's diffusion overlaps with the beginning of another, or the degree to which multiple templates diffuse at different but overlapping times. The dotted lines below represent the traditional breaks in history implied by traditional narratives. The image is intended to be illustrative; heights, widths, and overlap are approximations.

A diffusion-based account of prison history provides a more accurate depiction, one that is not easily represented linearly or tabularly. For illustrative purposes, Figure 1 offers a visual representation of developments from the late 1780s to the 1930s. (Although the figure uses approximations, I present data for two periods in the penultimate section.) The lower portion depicts a traditional visual representation of this period, wherein history is periodized into neat, self-contained periods associated with the dominant legal templates. The upper portion depicts a more empirically accurate representation of the same timeline: here, history is viewed as a series of different-sized, overlapping arcs associated with the dominant legal templates' adoption across different jurisdictions. By emphasizing templates' overlapping diffusion, this visual representation avoids traditional periodizations bookended by new templates' innovation.

While continuing to understand the process of innovation is essential, examining diffusion is important for at least three reasons. First, it immediately builds process into our analysis, which expands discussions of time from a moment (of innovation) to a duration (of diffusion). No longer a series of binary variables, time becomes a continuous variable. Rather than staccato periods, we see a series of diffusion curves, where diffusion curves are simply cumulative counts of a template's adoption (see Figure 2). Second, recognizing that diffusion is incomplete for a large portion of the period associated with a template challenges descriptions of history that imply periods of homogenous punishment once a new technology was immediately and uniformly adopted. Finally, examining diffusion complicates existing narratives about why jurisdictions adopt legal templates, even revealing different motivations driving adoption at different times within a single period.

Figure 2. A new view of penal change: hypothetical diffusion curves for three legal templates. This figure is for illustrative purposes: Templates A, B, and C could be, for example, proto-prisons, modern prisons, and adult reformatories, respectively; drop-style hangings, electrocution, and lethal injection, respectively; or individual law offices, the Cravath style law firm, or mega law firms, respectively. The overlap of the diffusion curves for Templates A and B illustrates how some templates begin to diffuse while previous templates are still diffusing, but their rate of diffusion is slowing. By contrast, Template C's diffusion curve begins once Template B's ends. In an empirical example, the height and width of each diffusion curve would vary depending on the duration of the diffusion and how widely the template is adopted. The categories of innovators, early adopters, late adopters, and laggards are borrowed from Rogers (2003). The dashed line represents an approximate point in the diffusion curve at which each template institutionalizes.

Duration

Although the speed of diffusion varies by template, several examples will illustrate how diffusion is frequently a long, drawn-out process. The diffusion of adult reformatories—prisons for first-time, redeemable young-adult offenders—in the late nineteenth century had a slow start: New York implemented the template in 1876, which was quickly copied by Michigan in 1877, but then more slowly by Massachusetts (1884), Pennsylvania and Minnesota (1889), another six states by 1898, and another eight by 1916, never reaching a majority of states despite 40 years of diffusion (Pisciotta 1994: 87). The diffusion of probation during the same period was even slower: Massachusetts authorized adult probation in 1878, followed nearly 20 years later by Missouri (1897), which was followed by 10 states over the next 10 years, followed by another 15 states over the next 10 years, and so on until 39 states had adopted adult probation by the late 1930s (Calahan, 1986: 173). In a more recent example, the pace of diffusion was much faster, but it still highlights the role of time. Nine states opened a supermax between 1986 and 1989, another 21 between 1990 and 1999, and another 12 between 2000 and 2005 (Reiter 2012a: 278–280).⁸ The typically prolonged process of diffusion adds new dimensions to our analyses of formal penal change.

Periodization

When a template's diffusion is relatively slow, it is misleading to associate that template with a particular decade or two. For example, adult reformatories are associated with the post–Civil War wave of reform and the brief return to rehabilitation in the 1860s and 1870s, but more than half of these reformatories were authorized after 1890. Likewise, probation is often attributed to the Progressive Era (roughly 1890 to the 1920s), even though probation emerged before that period began and continued to diffuse after the period was over. In both cases, the diffusion was never complete; indeed, for most of each period associated with reformatories and probation, states that adopted these templates were in the minority. Periodization becomes more problematic when examining competing templates. In the 1970s, then-laggard Arizona embraced correctional or rehabilitative templates around the same time that California, an innovator or early adopter of those templates, was openly rejecting them in favor of harsher templates (Allen 1981; Lynch 2010). Thus, as a descriptive matter, linear periodizations understate the messy, overlapping reality in which one template's diffusion begins while another template's still-incomplete diffusion is subsiding.

Motivations

The long, drawn-out, messy process of diffusion complicates past analyses of penal change that emphasize the social context surrounding the particular year or decade in which change began. Such accounts have long implied a certain cultural, political, or economic zeitgeist that made the new technology acceptable and even popular. While templates do not diffuse unless there is some connection to the relevant social world, some account that resonates with the adopting population, whether for cultural, religious, political, or educational reasons (Rogers 2003; Strang and Meyer 1993), the diffusion process suggests that such zeitgeist moments exist only for a handful of innovative or early adopting jurisdictions. In many cases, that moment has long since passed by the time the remaining jurisdictions adopt the template; to the extent that some cultural resonance aids late adoption, that resonance is likely to be weaker, or qualitatively different, than it was for innovators and early adopters. Thus, widespread adoption may not signify the sort of widespread philosophical shift to which a template's emergence is often attributed and may not even reflect the dominant philosophy at the time of adoption.

A Theory of Formal Penal Change

Any theory of formal penal change must therefore engage with a template's diffusion, not just its innovation. Combining the insights of neo-institutional theory and the law and society literature, and empirical examples from the punishment and society literature, I derive such a theory of formal penal change. Tracing the arc of a template's diffusion, the theory addresses why templates are authorized (rather than where they come from), how they impact punishment in practice, and how they change cognitive expectations about punishment. Under this theory, the key variable for understanding the significance of formal penal change is whether its affiliated legal template becomes institutionalized, which I argue causes widespread diffusion.

Legal Templates' Early and Late Adopters: Excavating the Diverse Reasons for Penal Change

Legal templates develop in jurisdictions that experience a particular change in circumstances that poses new challenges or (perceived) problems (e.g., apparent crime wave, increased immigration, changing political alliances, new race relations, tighter labor market, etc.).⁹ Even though several jurisdictions may experience the new challenge around the same time, only a few innovative jurisdictions will have the resources or will to create a legal template to address the challenge.¹⁰ Other jurisdictions facing a similar challenge, but perhaps without the resources or will to innovate, are likely to adopt one of the newly available templates, rather than creating their own solutions, to address their (similar) challenges. For these innovative and early adopting jurisdictions, then, the template solves a practical problem; adoption is a technically rational decision (Scott 2003: 33).¹¹

As more jurisdictions adopt the new legal template, it becomes institutionalized—taken for granted or widely accepted, perhaps even as the primary form of punishment (for a particular type of offender or crime).¹² Even though these jurisdictions may only imperfectly implement the template, their formal adoption conveys their approval of the template; increasingly, as more jurisdictions adopt the template, the template is understood to be consistent with society's norms and values (Berger and Luckmann 1967; Scott 1995: 59). Over time, it becomes the expected approach to punishment. Indeed, as a template becomes institutionalized, debate surrounding the template is likely to subside and the rate of adoption to increase (before eventually plateauing) (Tolbert and Zucker 1983: 23–24).¹³

The template's institutionalization can drive hold-out jurisdictions to adopt the template, thereby becoming late adopters or (for the last in line) laggards.¹⁴ These jurisdictions may not have experienced the same change in circumstances as the innovative and early adopter jurisdictions and thus have no technical need for the template. However, once the template is institutionalized, these hold-out jurisdictions begin to experience tremendous pressure to adopt it—what is now the expected form of punishment—or face challenges to their legitimacy (Tolbert and Zucker 1983: 26). For example, a police department in a town with no apparent gang problem will face, and accede to, community pressure to adopt a specialized gang unit (Katz 2001). Likewise, school districts with no drug problem will institute drug testing when drugs are considered a national problem; those districts that abstain become vulnerable to criticism that they are not doing their job to protect school children (Simon 2007, chapter 7). Late-adopting jurisdictions thus adopt the template to strengthen or protect their own legitimacy, rather than for reasons of technical rationality or efficiency (Meyer and Rowan 1977).

Legal Templates as Symbolic Structures: Revisiting the Limited Impact of Penal Change

In the end, many jurisdictions will have adopted the institutionalized template not because it is the superior response to some technical problem (whatever drove early adopters), but because it covers the jurisdiction in a mantle of legitimacy and provides conformity with expectations about what punishment should look like. However, such jurisdictions also have their own technical goals, which may or may not be stated openly, such as reducing crime, controlling visible disorder, or maintaining class- or race-based hierarchies. Importantly, institutionalized templates are not necessarily the most efficient means of achieving these technical goals. Consequently, a jurisdiction that adopts a template primarily for legitimacy reasons is likely to find the template a poor fit with local technical goals (Meyer and Rowan 1977).

These legitimacy-motivated jurisdictions must retain the template and pursue their technical goals. Importantly, templates' legitimizing power exists at the level of the visible—formal adoption: jurisdictions can satisfy their legitimacy needs by authorizing a template even if, privately, the template has little impact on actual practice. Indeed, jurisdictions will find ways to ensure the new template does not interfere with the ongoing pursuit of their technical goals. Specifically, they will “maintain gaps between their formal structures and their ongoing work activities” to pursue particular technical goals (Meyer and Rowan 1977: 341). This gap between theory and practice—between the visible adoption of legal templates and the invisible meting out of punishment—allows local actors, privately, to ignore the legal template or to adapt it to their own local needs or preferred ways of doing punishment (Meyer and Rowan 1977).¹⁵

For example, COMPSTAT, an approach to policing using statistics and computer software to guide decisions about allocating manpower and managing workloads, was created to address the specific crime problems unique to New York City. Deemed a success, it was adopted in other cities, including Lowell, Minneapolis, and Newark, each of which had different (and declining) crime challenges to which COMPSTAT was not well suited. Adopting a new protocol developed by the prestigious NYPD, however, imbued these other police departments with legitimacy. When COMPSTAT proved more cumbersome than helpful, these police departments largely reverted to their old routines and workloads while maintaining their visible commitment to COMPSTAT (Willis et al. 2007).

Jurisdictions may also alter the template behind the scenes and use it to their advantage. For example, a prison may adopt an inmate grievance procedure to satisfy juristic expectations and derive legitimacy from conforming to expectations about what constitutes fair treatment of prisoners. However, the procedure may be largely symbolic, rarely requiring changes to prisoners' treatment. Once in place, the template may be “co-opted” and used in ways unintended by the court rulings that initially mandated them, as when inmate grievance procedures are used to “cool out” frustrated prisoners or better surveil the prisoner population rather than address rights violations (Bordt and Musheno 1988; Calavita and Jenness 2015). So long as the template's imperfect implementation remains invisible, the jurisdiction's claim to employ the template continues to provide the jurisdiction with legitimacy.

Legal Template's Constitutive Power: Recognizing the Long-Term Consequences of Formal Penal Change

Despite its adulteration in practice, the institutionalized template remains powerful. Once a template becomes institutionalized, it becomes part of a society's expectations about what punishment is or looks like. Underlying the template's institutionalization is a general sense of acceptance or approval. This acceptance may have moral (it is proper) and pragmatic (it is best) dimensions, but importantly, it also has a cognitive-cultural dimension (it is how we do things) (Scott 1995; Suchman 1995). The template's growing acceptability or seemingly natural, primeval status starts to construct social understandings of punishment. Over time, these cognitive-cultural expectations become “prescriptive”—they shape new norms about what current and future punishments should look like, be like, or do (see Scott 1995: 35). Put differently, an institutionalized legal template becomes a “common frame of reference” when actors think of punishment (Scott 1995: 45). The template becomes the recognized substance of “the way we [punish]”; punishing differently becomes “inconceivable” (Scott 1995: 57). In this sense, legal templates become normative. I refer to this dynamic as legal templates' constitutive power (Ewick and Silbey 1998; Gordon 1984; Hirsch and Lazarus-Black 1994; Starr and Collier 1989).

This constitutive power has significant consequences for penal stasis and change. First, even if a legal template is ushered in by social elites whose views are not representative, as often occurs (Goodman et al. 2017), it can gain popular support by becoming institutionalized. Two extreme examples illustrate the point. The prisons emerging in the late eighteenth century received substantial opposition from prisoners, their family members, and even prison staff (Ignatieff 1978; McLennan 2008). Within 50 years, however, prisons were fully secure in the penal landscape—debates focused on the internal designs and routines of prisons, but there were no serious discussions about abolition. Today, while there are some abolitionists, most American citizens cannot imagine a criminal justice system that does not rely on the prison despite widespread agreement that the prison system needs substantial change (e.g., Gottschalk 2014). Likewise, France abolished its death penalty in 1981 by political mandate and against significant public opposition. Today, however, most French people oppose the death penalty and its return (Zimring 2006). The causal process by which public opinion changes is complicated and mediated by many factors, but I suggest that institutionalization contributes to this change: once a legal template is widely adopted (especially at the behest of social and political elites), it conveys perceptions of legitimate penal behavior that likely shape some significant part of public opinion about that punishment.¹⁶

Second, institutionalized legal templates provide the contours of future punishments. Once a template is institutionalized, reformers, legislators, penal administrators, or interest groups are free to tinker with or elaborate the template (Grattet et al. 1998). So long as they retain the underlying template in some recognizable form, jurisdictions can customize the template for their particular (local) needs and retain their legitimacy. For example, once early statutes penalizing hate crimes had institutionalized the general template of hate crime statutes (e.g., establishing the category of hate crime), later versions of these statutes could include a greater range of crimes and categories of protected persons. This sort of open “differentiation” is only possible once the original template is institutionalized (Grattet et al. 1998: 303).¹⁷ The initial template becomes so well ensconced in the penal landscape that, even when the template seems a poor fit for the challenges at hand, jurisdictions create new versions of the template rather than seeking altogether different templates.

Legal templates' constitutive power determines their larger significance by causing lasting change. Indeed, templates' constitutive power may explain why, “[o]nce in place, the infrastructure of penality attains a certain resilience and inertia, and major changes in that apparatus tend to occur slowly and against a great deal of resistance,” even when goals and actual practice change more quickly (Garland 2003: 64).

Applying the Theory: A Pair of Illustrative Examples

Thus far, this article has offered diverse examples from the punishment literature to illustrate the theory's generalizability. In this section, I provide a more systematic application by synthesizing my own primary research examining two overlapping periods of penal change.¹⁸ Traditional prison histories associate these periods with the rise of the proto-prison or penitentiary house (1790s–1810s) and of the modern prison or factory-style prison (1820s–1860s).¹⁹ Instead, I discuss multiple legal templates for incarceration that developed in each period, but I focus on the process by which certain templates became dominant, including the reasons for their diffusion and institutionalization, their relationship to actual practice, and their constitutive power. In each case, I defer discussions about the larger social context and its relationship to a particular template's emergence, including what sorts of values, theories, or beliefs shaped the templates' contours; these contextual influences have been amply discussed by extant literature. Instead, I focus on the components contributed by the legal templates theory. Importantly, these components do not conflict with the existing literature but instead help explain the course of penal change after it has begun, especially after a particular template has been designed. This post-innovation phase of penal change has been less explicitly addressed by extant prison histories.

Proto-Prisons, 1785–1822

Overlapping and Varied Diffusion

Following the American Revolution, penal reformers and some state governments aggressively sought alternatives to their traditional capital and corporal punishments. Proto-prisons—state-run facilities that incarcerated convicted criminals from across the state at hard labor for a year or more—soon became the dominant template. However, proto-prisons were not the first nor the only template available. Although not originally places of punishment, colonial jails offered one template: these administrative adjuncts were locally run facilities that confined a mix of vagrants, debtors, witnesses, accused criminals awaiting trial, and convicted offenders awaiting their (noncarceral) punishments. Some states began incarcerating offenders for short periods in their local jails to perform hard labor; other states turned to public labor, with nightly confinement in jail. One particularly distinctive version of public labor was Pennsylvania's “wheelbarrow law,” under which convicted felons—affixed to a ball and chain and clothed in distinctive, degrading garb and hairstyles—would clean and repair the streets (Meranze 1996). Although these other templates were copied by a few states, they proved less popular than proto-prisons and thus had little direct impact on subsequent penal history (Rubin 2018).

Innovation

Initially, several states created different versions of the proto-prison. The first proto-prison was established on an island military fort in Boston Harbor (1785); convicted offenders from across the state were sent to be treated like enlisted men and kept under military discipline at the Castle Island prison. The second proto-prison, Newgate, was housed at a Connecticut coal mine (1790) and was thought to be inescapable: the state's prisoners would descend by ladder to work in underground caverns. The third proto-prison, a restructured late-colonial jail on Walnut Street in Philadelphia (1790–1794), still held vagrants, debtors, and others but kept them separate from the convicted criminals, who were themselves segregated by age (or level of criminality) and sex; these convicts were set to hard labor and punished with solitary confinement if they misbehaved (Rubin 2018). None of the proto-prison's three versions was particularly well developed, but one quickly became popular—so popular that, as often happens, the version became synonymous with the template and the other versions disappeared from the penal landscape.

Early Adopters

Walnut Street Prison was centrally located in the nation's cultural (and previously political) capital and benefited from numerous visitors and vocal proponents who publicized its “penitentiary house” design and detailed its inner workings. These descriptions were widely circulated, encouraging reformers and statesmen to visit the prison. The Walnut Street version of the proto-prison template was copied first by New York and Virginia in 1796, followed by New Jersey in 1797 and Kentucky in 1798. For these early adopters, the proto-prison template solved a practical problem that had recently developed: the need to restrict their use of capital and corporal punishment. Pennsylvanian reformers and legislators had developed Walnut Street Prison as an alternative to corporal and capital punishment; the Pennsylvania legislature ultimately limited capital punishment to first-degree murder in 1794, the same year Walnut Street officially became a state prison. When out-of-state reformers visited, they simultaneously recommended that their own states should abolish or restrict capital punishment and adopt a proto-prison along the lines of Walnut Street. Indeed, the first states to adopt the Walnut Street version—New York, Virginia, New Jersey, and Kentucky—likewise limited their capital crimes to murder in the same year (or the year before) they authorized their proto-prison (Rubin 2016).

Late Adopters

Between 1800 and 1822, another 12 states authorized a proto-prison modeled on Walnut Street (Rubin 2018). The reasons behind these various states' adoptions were more variable than the early adopters' reasons. Indeed, these states adopted a proto-prison even though they retained a wide range of capital offenses (at the time of adoption); several states adopted a proto-prison years before legally restricting capital punishment, while others expanded their capital statutes (Rubin 2016). Thus, most of these proto-prisons were not adopted as solutions to the same technical problems that drove early adopters. Moreover, diffusion of the Walnut Street version after about 1800 cannot be attributed to its technical success: as we shall see, by that time, Philadelphia's reformers were already disappointed with their prison and publicly sought alterations (Rubin N.D.). Instead, by 1800, proto-prisons following the Walnut Street version were sufficiently widespread to become taken for granted as a proper means of punishment that late-adopter states continued to erect proto-prisons modeled on Walnut Street despite its apparent problems. Indeed, no new versions of the proto-prison emerged in this period. Rather than creating an alternative version of the proto-prison (or alternative templates to the proto-prison), states borrowed the version that had become increasingly common and expected.

Relationship to Practice

The proto-prison offered a template for the newly popular idea to incarcerate prisoners over the long term as punishment, but there were significant gaps between theory and practice. Across the country, most proto-prisons, including Walnut Street, did not function as they were publicly described. Many proponents emphasized the disciplinary effects of solitary confinement and labor and the palliative role of segregation; in practice, there were too few solitary cells to be used for any meaningful period (according to reformers and prison managers alike) and overcrowding and financial difficulties prevented regular labor or proper segregation. Guards and prisoners also widened the gap by subverting the rules; by the 1810s, an onslaught of fires, riots, escapes, and an apparent increase in the crime rate (attributed to the proto-prisons) graphically illustrated the proto-prison's failure to perform as expected (e.g., McLennan 2008; Meranze 1996; Rubin 2015).

Constitutive Power

The proto-prisons' ongoing failure precipitated a new wave of reform (Rubin 2015). Initially, reformers and statesmen discussed alternatives to their failing proto-prisons. Some suggested reviving and expanding the death penalty while others suggested abolishing the prison. However, most solutions kept the proto-prison at the center. For example, some states authorized prison managers to whip misbehaving prisoners or passed statutes that made arson in prison a capital offense—both of these alternatives built onto the theory the proto-prison provided (Rubin 2016, 2020). As we shall see, however, the most popular strategy was simply to build bigger, stronger, more rigorous versions of the proto-prison, which was, by then, institutionalized: despite its failure, the perceived solution was more of the same. Notably, initial experiments with what would become a new template drew on some of Walnut Street's most publicized, but least utilized, dimensions—especially solitary confinement—as reformers, prison managers, and legislators sought to expand and improve the proto-prison template. The institutionalized template constrained the realm of the possible: the proto-prison was so taken for granted that meaningful alternatives could not replace or displace it.

Modern Prisons, 1816–1870s+

Overlapping and Varied Diffusion

Out of the ashes of proto-prison arsons and riots, modern prisons emerged as the next dominant template. Whereas proto-prisons' daily routines were less structured and provided prisoners a fair amount of freedom and socialization, the new modern prisons sought to expand prison managers' control through tighter schedules and stricter segregation, whether social (silence) or physical (solitary). Notably, the first modern prisons opened at the tail end of the proto-prison's diffusion, while frontier states and established but laggard states were still authorizing their first proto-prisons (see Figure 3). Instead, the new wave of reform was again driven by states with emergent technical needs: Pennsylvania and New York housed some of the largest prison populations in the oldest facilities, which suffered the most extensive and frequent disorder in the 1810s. Both states authorized larger, stronger, more tightly controlled facilities in separate parts of their states (upstate New York in 1816 and western Pennsylvania in 1818). Initially, these new prisons were not understood as new templates or challengers to the proto-prison. Instead, the new facilities were modeled on the old template, but their larger size and rural location were expected to alleviate the overcrowding in the states' aging proto-prisons and reduce the expense of transporting prisoners from more rural parts of each state. As the existing proto-prisons deteriorated, however, these states revised their plans for the prisons under construction to address these problems, creating new templates in the process—the first modern prisons (Rubin 2020).

Figure 3. The diffusion of proto-prisons and modern prisons in America, 1785–1865. The vertical lines indicate the traditional period breaks. For consistency across periods, I have used date of opening, rather than the authorization date. Additional modern prisons adopted after 1865 are not shown. Source: Rubin (2015, 2016, ND). [Color figure can be viewed at wileyonlinelibrary.com]

Innovation

Both states, in quick succession, created several versions of the modern prison. The first featured total solitary confinement. Pennsylvania's 1818 statute ordered all prisoners at the new prison to be kept in solitary confinement for the duration of their sentence. Likewise, in 1819, on the heels of further rioting, New York's legislature authorized an additional cellblock at Auburn State Prison, its new upstate prison under construction, to keep prisoners in solitary confinement when it opened in 1821. After only a few years, Auburn's long-term solitary confinement yielded highly publicized negative results—disease, insanity, suicide, and premature death—prompting authorities to end the system and replace it with another version. Since Auburn opened, another group of prisoners had been kept in solitary confinement only at night; clad in striped uniforms, they marched from their cells in lockstep to work silently in factory-like settings during the day, an approach that became known as the “Auburn System” (or “Silent System”). Meanwhile, Pennsylvania, facing other challenges with long-term solitary confinement, altered its plan by adding labor and redesigning its prisons to include larger cells and private yards. Additionally, it prescribed visitation from prison officials and local penal reformers who would train, educate, and mentor the prisoners. The approach was called the “Pennsylvania System” (or “Separate System”) (Rubin 2015).

Early Adopters

The modern prison template, especially the Auburn System version, spread rapidly. Within 10 years of its first appearance, modern prisons following the Auburn System had been authorized in Massachusetts, Maryland, Kentucky, Tennessee, Maine, and Vermont, as well as another modern prison in New York. Another few years added a prison in the District of Columbia, as well as Georgia, New Hampshire, Ohio, Illinois, and Louisiana. The Pennsylvania System diffused more slowly and less widely: in addition to two modern prisons in Pennsylvania, both New Jersey and Rhode Island built prisons following the Pennsylvania System. Notably, the vast majority of early adopter states had previously built a proto-prison: the modern prisons were designed and adopted to resolve problems that had become endemic in those older facilities (Rubin 2015).

Late Adopters

By 1835, a plurality of states had authorized or built a modern prison, the vast majority following the Auburn System. At this point, the modern prison template (virtually synonymous with the Auburn System version) had become so widespread and taken for granted, that it created its own gravitational force. Those southern states continuing to use capital and corporal punishments (and slavery) over incarceration appeared aberrant against “enlightened,” “modern” trends. For them, the modern prison was an instant source of legitimacy. New states on the frontier also turned to the new template to illustrate their bonafides as states. Intensifying this need to conform to trends, penal reformers made nationally circulated reports criticizing states that resisted the Auburn System, their recommended version of a now-mandatory template. Indeed, Rhode Island and New Jersey abandoned the Pennsylvania System in favor of the Auburn System. By the Civil War, all but four of the 34 states had built or authorized a modern prison and only Pennsylvania retained the Pennsylvania System (Rubin 2015).

Relationship to Practice

The modern prison, whether the Auburn System or Pennsylvania System versions, was a template for incarceration that, like the Walnut Street–style proto-prison, guided actual practice intermittently. In Auburn-style modern prisons, the primacy of labor and the influence of entrepreneurial contractors often superseded the System's rules, including its emphasis on silence (McLennan 2008). Variations in practice also emerged early in Kentucky and Maryland, although commentators still considered these prisons representatives of the Auburn System (Rubin 2015: 373, n. 3). Similarly, prison managers violated the Pennsylvania System's mandate to keep prisoners physically separate by double-celling prisoners and releasing them from their cells to work around the prison (Rubin 2020). The image of modern prisons as tightly controlled, well-ordered facilities—perpetuated by penal reformers and prison managers' public descriptions—was only occasionally true of individual facilities at certain times in their history. Instead, other realities—expenses, overcrowding, staffing, and profit motivations—quickly created cleavages between theory and practice (McLennan 2008; Rubin 2020).

Constitutive Power

Well after the Civil War, the modern prison template (now synonymous with the Auburn System version) continued to shape penal innovations even after its problems were apparent.²⁰ Northern prison populations surged immediately after the war, causing overcrowding. Rather than seeking alternative forms of punishment to reduce the prison population and divert those prisoners that seemed a poor fit for prison, large states authorized second, third, or fourth prisons to take up the slack. These new prisons retained many features of the Auburn System, but were further specialized for target populations creating new templates in the process: states built women's prisons and adult reformatories for the “redeemable” prisoner, while older prisons were redesignated high-security prisons and reserved for the “worst” offenders (Pisciotta 1994; Rafter 1985; Rothman 1980). Each new template had its own specific programs and components, but they were variations on a theme; the institutionalized prison continued to limit the penal imagination (Rubin 2014).

In the South, even though the modern prisons had been destroyed during the war, their constitutive power remained. Southern states blended modern forms of incarceration—especially the Auburn System's practice of allowing private contractors to lease prisoners' labor—with another institutionalized practice, slavery, to create convict leasing. Under this new hybridized template, “prisoners” (mostly former slaves wearing Auburn's black-and-white striped uniform) were transplanted to fields, swamps, and forests and housed in mobile cages and bunkers until states eventually created permanent plantation-style prisons—a new template that merged chattel slavery and the Auburn-style modern prison (e.g., Oshinsky 1997).

Indeed, the modern prison's constitutive power continued later in the century when reformers, dissatisfied with incarceration, turned to community supervision, a new mode of punishment, and its two templates, parole and probation. Although reformers saw prisons as criminogenic, they designed parole and probation as adjunct punishments for specific populations of prisoners—most of whom would remain incarcerated. Moreover, while probation was billed explicitly as an alternative to prison, those who failed their probation were sent to prison, swelling the incarceration rate in the process (Cohen 1979; Rothman 1980). These alternative punishments, designed in response to the prison's failure, built prison into their new recipe for punishment (Rubin 2014).

Thus, in the decades after the war, a diverse range of states created new templates, but these templates again drew heavily on existing templates: states simply customized institutionalized templates to address various problems associated with different prisoner populations and new postwar circumstances (Rubin 2014). Despite each prison templates' repeated failures and incongruity with new circumstances, their constitutive power survived, shaping the next generation of templates.

Discussion and Conclusion

Beginning with Emile Durkheim, scholars have placed great weight on changes in punishment's visible landscape, using the rise and fall of new penal technologies as signifiers of widespread penal change, thought to reflect broader changes in society at large. Recent research, however, has illustrated multiple problems with such analyses of penal change. With the concept of legal templates, or structural models of punishment available for authorization and replication, this article has forged a middle ground. I have argued that new legal templates are important signifiers of penal change, but not in the ways traditionally associated with new forms of punishment. Legal templates rarely revolutionize the practice of punishment in the ways imagined, nor are they the result of great ideological change; but institutionalized templates are signifiers of change—of changes to society's expectations about punishment.

Importantly, the legal templates theory resolves traditional approaches' three limitations. First, this theory provides a new way of thinking about templates' temporal and spatial dominance by emphasizing diffusion—frequently a lengthy and incomplete process—rather than implying immediate, universal acceptance lasting throughout the template's existence. The adoption of proto-prisons and modern prisons was much slower than traditional zeitgeist accounts imply. Second, understanding the role of loose coupling directly incorporates and explains the perennial gap between theory and practice as the need to balance local technical needs with external legitimacy demands. For example, prisons officially following the Auburn System violated its emphasis on silence to maximize profit but formally maintained the system for external legitimacy. Finally, examining the institutionalization process explains how templates promoted by a small group of elites become popular. Prisons pioneered by northern penal reformers eventually spread throughout the entire country, despite distinct local needs and preferences, eventually becoming taken for granted.

More generally, legal templates' constitutive power provides the missing basis for the significance of formal penal change. This theory ultimately shows that, regardless of new legal templates' impact on actual, on-the-ground penal practice, the institutionalization of these templates has long-term consequences for ideas about punishment. Even when templates failed, next-generation reformers retained significant components for future templates. In effect, legal templates provide the boundaries within which future penal trends emerge, casting a shadow over future discussions about punishment and penal change.

Limitations and Future Research

Although the legal templates theory provides a powerful tool kit for understanding penal change, it has two major limitations. First, unlike prior accounts of penal change, this theory takes no stand on the underlying cause, the spark that initiates change. This theory refers to technical-rational needs—meaning a jurisdiction's actual goals, whether crime control, race/class control, or some other desired end—but not where these needs come from. Technical-rational needs may appear for many reasons—new ideas about crime and criminals, significant overcrowding, the abolition of slavery, the (perceived) failure of Keynesian economics, a dramatic increase in immigration, court decisions ruling prior practices unconstitutional—that will be specific to the time and place. The templates theory thus makes no general statements about the arrival of new technical-rational needs; consequently, it works best in combination with other accounts that do address the underlying causes or sparks of penal change, such as Goodman et al.'s (2017) agonistic perspective.

The agonistic perspective argues that macro-level shifts in society alter the relative power and influence of different groups, which are then better (or less) able to impose changes that reflect their particular penal agenda. We can also anticipate that macro-level shifts create new technical-rational needs; new or existing groups may propose adopting a particular template to resolve these needs. The groups' influence may help the template spread to other early adopter jurisdictions. For example, the group may circulate myths about the template's utility, which help construct the template as a beneficial policy or practice to adopt, imposing greater pressure on those jurisdictions that fail to adopt it (DiMaggio and Powell 1983; Rubin 2015). Finally, the template's growing institutionalization may help explain its adoption in those jurisdictions where ascendant groups are not influential.

Second, this theory applies only to formal penal change—especially the authorization (or abolition) of particular legal templates—and its consequences. Although this theory incorporates the gap between expectations and practice, it emphasizes punishment's structures rather than trends in their usage. Indeed, while this theory allows for frontline workers and other discretionary agents (e.g., police, parole officers, judges, juries), it does not discuss the informal (or cultural) bases of their decisionmaking, such as individual factors like racism, religion, or socioeconomic status or larger contextual factors. To understand actual practices, such as sentencing decisions or overall incarceration rates, we must turn to multilevel analyses that incorporate both new templates and local actors' use of them (Campbell et al. 2015; Campbell and Schoenfeld 2013).

The templates theory extends beyond punishment to its parent institution, the law. In the criminal law, we can analyze penal codes, the distinction between first- and second-degree murder, repeat-offender laws, namesake laws, sentencing guidelines, and crime victims' bills of rights as templates. More broadly, some of the most significant legal phenomena—jury trials (Hans 2017), alternative dispute resolution (Edelman and Suchman 1999), EEO/AA policies (Edelman 1992), problem-solving courts (Mirchandani 2005), constitutions (Law and Versteeg 2011), the Cravath model for law firms (Nelson 1981), and mega law firms (Liu and Wu 2016)—are themselves templates that diffuse, institutionalize, and shape conceptions of the law, its institutions, and its practices, even while functioning differently than anticipated in practice. Indeed, thinking of legal templates—like alternative dispute resolution or EEO/AA policies—as formal structures loosely coupled to actual practice and recognizing their constitutive power is already well established within law and organizations studies (Edelman 1992, 2016). Under legal endogeneity theory, “the meaning of law is shaped by widely accepted ideas within the social arena that law seeks to regulate,” thereby blunting the law's intended effect (Edelman 2016: 12). Under the legal templates theory, however, the templates themselves begin to shape wider social expectations about the law, constraining future efforts to change even after the templates have failed. Moving forward, the insights of these two theories can be combined and extended to examine other legal templates.

Conclusion

The law may be a tool with limited potential to create social change, but it also affects society unexpectedly and significantly through its constitutive power. The adoption of new legal templates may not shape actual practice in the way policymakers plan. Once the template is institutionalized, however, its taken-for-granted status may shape future expectations about what law and punishment should look like and thereby constrain citizens' ability to imagine new ways of organizing society or punishing its citizens. In the process, the template takes on a life of its own, far outlasting its creators' intentions, by anchoring future legal and penal change. In this way, the advent and institutionalization of new legal templates is far more powerful, and powerful in different ways, than previously understood.